ANN for automatic shaping of the multi-contact surface electrodes

We developed a method for automatic selection of the contacts within the multi-contact surface electrode that leads to the optimal grasping in individuals with tetraplegia. The method uses Artificial Neural Network that estimated the optimal grasp based on the joint angle excursions. Our method consists of initial joint angle measurement while activating each field of multi-contact surface electrode, artificial neural network training with acquired data and optimal electrode calculations using trained neural network.